Research on the Application of University Library Management System Based on Hash-Based Technique and K-Nearest Neighbours-Collaborative Filtering

In recent years, Big data technology increased tremendously that helped libraries in improving their services by providing insights into user behaviour and needs. These technologies improved personalized recommendations and enhanced digital collections in university libraries but managing the university library system also important. The existing methods for library management system failed in recommending useful books to users that resulted with decreased transactions of library. Therefore, this paper proposes Hash-Based Technique and K-Nearest NeighboursCollaborative Filtering (HBT-KNN-Collaborative Filtering) method for application of university library management system based on book recommendations. The proposed method employed on open library dataset that incorporates university library books data. Then, this data is pre-processed by using min-max scaling normalization for scaling all data values into similar ranges. These pre-processed data is mined using HBT for generating association rules. Finally, recommendation of books is done by using KNN-Collaborative Filtering which analysed K-neighbours for efficient high rated and frequently borrowed books recommendation. The proposed HBT-KNNCollaborative Filtering achieved better results in terms of support (16.28%) and confidence (69.58%) Score when compared with existing apriori algorithm.